I'll try to answer all mails at once :-)) - query is running fully in RAM, no I/O, no network, only CPU time - looping 100 times the same query gives 132ms total time (~1.32ms per query), while it's 44ms on InnoDB (~0.44ms per query) - disabling seq scan forcing a planner to use an index scan, and finally it worse as gives 1.53ms per query.. - prepare the query helps: prepare statement takes 16ms, but execute runs in 0.98ms = which make me think it's not only a planner overhead... And it's still 2 times lower vs 0.44ms. Also, generally prepare cannot be used in this test case as we suppose any query may be of any kind (even if it's not always true :-)) - char or varchar should be used here because the reference code is supposed to accept any characters (alphanumeric) - it also reminds me that probably there are some extra CPU time due locale setting - but all my "lc_*" variables are set to "C"... Rgds, -Dimitri On 5/6/09, Merlin Moncure <mmoncure@xxxxxxxxx> wrote: > On Wed, May 6, 2009 at 7:46 AM, Merlin Moncure <mmoncure@xxxxxxxxx> wrote: >> prepare history_stat(char(10) as > > typo: > prepare history_stat(char(10)) as > -- Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-performance
